Project Summary: The main objective of the proposed research is to understand how short-term memory (STM) processes interact with sentence production, both in normal and in brain-damaged individuals. To this end, the proposed experiments will examine several sentence production phenomena in aphasic patients with specific types of STM deficits, and in neurologically normal control subjects. One set of proposed experiments will investigate speakers' tendency to choose syntactic structures allowing for early mention of both accessible and of given information (accessibility effects and given-new ordering), to see what kinds of memory representations underlie these effects. These effects will be assessed both in tightly controlled experimental methodologies and in more naturalistic production tasks. This will not only extend the validity of our findings, but also will provide a rich source of sentence production data to generate future hypotheses. The second set of proposed experiments will investigate speakers' tendency to repeat syntactic structures (syntactic priming), and investigate the role of inhibition in speakers' choices between syntactic alternatives. In both sets of experiments, the goal is to better understand the mechanisms underlying these sentence production phenomena by examining how patient groups dissociate from control groups, thus illuminating the role played by STM in sentence production. The results of these studies should inform both theories of sentence production and of STM. Finally, after receiving training in lesion-deficit analysis and functional MRI, further experiments will be designed and conducted during the final year of support. These studies should further illuminate the relationship between STM, inhibition, and sentence production. Relevance: Problems with sentence production and with short-term memory (STM) are associated with a variety of disorders (including aphasia and Alzheimer's disease). A better understanding of how sentence production relies on STM should be helpful not only for theories of language production and of memory, but also should lead to a better understanding of exactly what underlies the language deficits in brain-damaged patients. This, in turn, is likely to aid in the development of effective therapies. [unreadable] [unreadable] [unreadable]